Valentina Lauria. A thesis submitted to Plymouth University in partial fulfilment for the degree of. DOCTOR OF PHILOSOPHY. School of Science and Technology.
This copy of the thesis has been supplied on condition that anyone who consults it is understood to recognise that its copyright rests with its author and that no quotation from the thesis and no information derived from it may be published without the author’s prior consent.
Impacts of climate change and fisheries on the Celtic Sea ecosystem
Valentina Lauria
A thesis submitted to Plymouth University in partial fulfilment for the degree of
DOCTOR OF PHILOSOPHY
School of Science and Technology Faculty of Marine Science and Engineering In collaboration with Natural England, Sir Alister Hardy Foundation for Ocean Science and the Centre for Environment, Fisheries and Aquaculture Science
April 2012
Abstract Impacts of climate change and fisheries on the Celtic Sea ecosystem Valentina Lauria Climate change and fisheries have affected marine environments worldwide leading to impacts on ecosystem structure and functioning. However there is clear evidence of spatial variability in the response of these impacts both within and among marine ecosystems. Although several studies have tried to explain the effect of these impacts on marine food webs, it is unclear how they interact, and how they may affect marine ecosystems remains an important unanswered question. This suggests the urgent need for multiple-trophic level and ecosystem-based management approaches to account for both fisheries and climate change impacts at ocean basins across the globe. Marine apex predators, such as seabirds, are vulnerable to the effects of both climate and fishing impacts, and can be used as reliable and sensitive bio-indicators of the status of the marine ecosystem. The Celtic Sea ecosystem is a productive shelf region in the Northeast Atlantic. It is characterized by high fish and invertebrate biodiversity. In addition, internationally important numbers of seabirds, such as Northern gannet Morus bassanus (L.), Manx shearwater Puffinus puffinus (B.), Common guillemot Uria aalge (P.) and Black-legged kittiwake Rissa tridactyla (L.), breed along the Celtic Sea coasts. In recent years, fisheries from across Europe have intensively exploited the Celtic Sea, leading to changes in stock structure. Moreover, the increase in annual average Sea Surface Temperature by 0.67 oC over the past two decades has altered the composition of plankton communities. These impacts, independently and in tandem, are likely to have had dramatic effects upon the
i
Celtic Sea food web emphasizing the need to enhance our understanding of this important marine ecosystem. In this thesis the effects of climate change and fisheries on the Celtic Sea pelagic food web are evaluated, in particular focussing on the response of seabird populations. This is in part because of recent declines in the breeding success of many seabird colonies in the northeast Atlantic, particularly around the North Sea. Long-term data across four trophic levels (phytoplankton, zooplankton, mid-trophic level fish and seabirds) and different modelling approaches are used to determine factors influencing seabird productivity at different geographical scales. First, I review the direct and indirect effects of climate change and fisheries upon marine ecosystems, as well as their impacts upon marine birds. Second, I use data collected during 1986-2007 from a single seabird colony, across four trophic levels, to investigate long-term direct and indirect climate effects. The results suggest only a weak climate signal in the Celtic Sea, and this is only evident between mid-trophic level fish and certain species of seabird. Third, a similar multi-trophic level approach across three nearby regions in the southwest UK (Irish Sea, Celtic Sea, and English Channel) reveal no evidence of a bottom-up signal during the period 1991-2007. These findings are in contrast with the nearby North Sea region, where a strong bottom-up effect was found to affect seabird populations, highlighting the importance of regional-based studies across multiple trophic levels. Finally, to provide a more complete picture of the Celtic Sea, and how it might respond to changes in fisheries management and climatic variation, I use the complex tropho-dynamic ecosystem model Ecopath with Ecosim. The main focus is on how seabird biomass changes in response to the application of different fisheries regimes likely to be implemented under forthcoming reforms to the Common Fisheries Policy (e.g. the application of quotas and discard bans), as well as future climate change scenarios, in order ii
to provide guideline support for resource management and seabird conservation in the Celtic Sea. The results suggest that some seabird guilds (gulls and some other scavengers) may be negatively affected by a reduction in discards, while other species (offshore divers) will benefit from a decrease in the fishing of pelagic fish species. Climate change is likely to have a negative impact across all trophic levels with a strong negative impact upon seabird populations. Therefore seabirds are likely to show species-specific responses to both climate variation (bottom-up effect) and changes in fishing practices, in particular our findings suggest that for some species climate may outweigh the fisheries impacts even when fisheries pressure is reduced by 50%. In summary, this study suggests that despite the generally negative impact of climate described for some regions in the Northeast Atlantic, the Celtic Sea ecosystem seems to be more resilient. However, both climate and fisheries and the interactions between these factors should be taken into account in the formulation of future management plans for the Celtic Sea ecosystem. The use of multiple-trophic level and ecosystem-based approaches over multiple spatial and temporal scales has helped to elucidate possible trophic mechanisms that are the response to future fishing and climate impacts in the Celtic Sea. The results of this study could have implications for both management plans and conservation policy.
iii
Table of contents i iv vi xv xvi
Abstract Table of contents List of figures and tables Acknowledgements Author’s declaration Chapter 1: General introduction 1.1 Marine ecosystem change 1.2 Theoretical considerations: seabirds as indicators of marine ecosystem change 1.3 Seabirds-Climate change effects 1.4 Seabirds-Fishery interactions 1.5 Conclusions 1.6 Aims of the thesis
1 4 5 8 11 11
Chapter 2: Influence of climate change and trophic coupling across four trophic levels in the Celtic Sea 14 15 17 30 34 36
Abstract 2.1 Introduction 2.2 Materials and Methods 2.3 Results 2.4 Discussion 2.5 Conclusions Chapter 3: Regional variability and climate change effects in the pelagic food-web of three marine ecosystems around Great Britain
39 40 43 53 60 63
Abstract 3.1 Introduction 3.2 Materials and Methods 3.3 Results 3.4 Discussion 3.5 Conclusions Chapter 4: Trophic relationships of seabirds in the Celtic Sea ecosystem: a view from an ECOPATH model
65 66 69 76 83 87
Abstract 4.1 Introduction 4.2 Materials and Methods 4.3 Results 4.4 Discussion 4.5 Conclusions iv
Chapter 5: Predicting fishing and climate effects on marine apex predators in the Celtic Sea using a tropho-dynamic simulation model Abstract 5.1 Introduction 5.2 Materials and Methods 5.3 Results 5.4 Discussion 5.5 Conclusions and Recommendations
88 89 92 98 105 109
Chapter 6: Discussion and Conclusions
111
APPENDIX 1 APPENDIX 2 APPENDIX 3 APPENDIX 4 References
118 123 129 179 200
v
List of figures and tables CHAPTER 2 Figure 2.1 The Celtic Sea, including oceanographic features (adapted from OSPAR, 2000) and the seabird colony investigated in this study (Skomer Island: 51 º 40N, 05 º 15W). The Irish Shelf Front occurs to the south and west of Ireland and exists all year-round. This front marks the boundary between waters of the shelf (often mixed vertically by the tide) and offshore North Atlantic waters. In addition there are two seasonal fronts systems which tend to develop during spring: the Celtic Sea front (dividing the Celtic Sea from the Irish Sea) and the Ushant Front, which develops from the coast of Brittany and extends to the western English Channel (dividing the Celtic Sea from the English Channel).
Table 2.1 Potential climate effects for four trophic levels in the Celtic Sea. Direct effects are manifest by correlations between climatic predictors and one of the ecological descriptors. Indirect effect links ecological descriptors to climate only through its effect on another trophic level i.e. via trophic coupling.
Figure 2.2 Variables used for model construction: Winter NAO index (a); Spring NAO index (b); Sea Surface Temperature (ºC) (c); Diatom abundance (d); small and large copepods biomass (mg wet weight) (e); Herring 0- and 1-group abundance (f); Blacklegged kittiwake productivity (number of fledged chicks per breeding pair, weighted for sample size) (g) and population count (h); Common guillemot productivity (number of fledged chicks per breeding pair, weighted for sample size) (i) and population count (j); Razorbill productivity (number of fledged chicks per breeding pair, weighted for sample size) (k) and population count (l); Atlantic puffin productivity (number of fledged chicks per breeding pair, weighted for sample size) (m) and population count (n). Fitted linear regressions indicate significant temporal trends. Table 2.3 Model selection to estimate factors influencing each trophic level. Only the best supported models are shown. AICc weight: Akaike‟s Information Criteria (corrected) weights, values range from 0 to 1, and high values indicate strong support for a given predictor; Er: Evidence ratio k: number of parameters in the model; R2: adjusted coefficient. vi
WNAO: winter North Atlantic Oscillation index; SNAO: spring North Atlantic Oscillation index; WSST: winter Sea Surface Temperature; 1lag-SSST: 1 year lagged spring Sea Surface Temperature; 2lag-SSST: 2 years lagged spring Sea Surface Temperature; her 0-g: herring 0-group; her-1g: herring 1-group; Significant relationships are highlighted in bold; variables that are not statistically significant but feature in the best model are also presented.
Table 2.4 Impact of climate variability across multiple trophic levels in the North Atlantic. SST: Sea Surface Temperature; NAO: North Atlantic Oscillation index.
CHAPTER 3 Figure 3.1 (a) Map of the study area, including frontal systems (adapted from OSPAR, 2004). The locations of kittiwake (Rissa tridactyla) colonies investigated in this study are also presented; 1: Great Ormes Head (53° 20N, 3° 51W), 2: Bardsey Island (52° 72N, 4° 77′W), 3: Skomer Island (51º 40N, 05º 15W), 4: Elegug Stacks (51º 60N, 04º 98W), 5: Dunmore East (51º 15N, 06º 99W), 6: Ram Head (51º 15N, 07º 70W), 7: Durlston Head (50º 54N, 02º 02W). (b) Map of the study area, showing the locations of all CPR samples for the three regions from March to June. In blues samples used for the Irish Sea (52ºN-55º, 7ºE-2ºW) spanning from 1991-2007, in green samples used for the Celtic Sea (50ºN-52º, 10ºE-5ºW) spanning from 1991-2007, in red sample used for the English Channel (49ºN51º, 10ºE-1ºW) spanning from 1991-2004.
Figure 3.2 Variables used for the models construction for each region: Sea Surface Temperature (ºC) (a-c); Diatom abundance (d-f); copepods biomass (mg wet weight) (g-i); fish larval abundance (j-l); Black-legged kittiwake productivity (number of fledged chicks per breeding pair, weighted for sample size) (m-p). The line indicates when temporal trends are present. Note the scales differ.
Table 3.1 Response variables and predictors used for the multiple regression models. For each response variable the full model is also given. WSST: Winter Sea Surface Temperature; 1Lag-SST: 1 year lagged annual Sea Surface Temperature; Kittiwake BS:
vii
Breeding Success expressed as the number of fledged chicks per breeding pair per year, weighted for sample size.
Table 3.2 Model selection to estimate factors influencing each trophic level for the Irish Sea. Only the best supported models are shown. AICc weight: Akaike‟s Information Criteria (corrected) weights, values range from 0 to 1, high values indicating strong support for a given predictor; k: number of parameters in the model; R2: adjusted coefficient. Significant relationships are highlighted in bold; variables that are not statistically significant but feature in the best model are also presented.
Table 3.3 Model selection to estimate factors influencing each trophic level for the Celtic Sea. Only the best supported models are shown. AICc weight: Akaike‟s Information Criteria (corrected) weights, values range from 0 to 1, high values indicating strong support for a given predictor; k: number of parameters in the model; R2: adjusted coefficient. Significant relationships are highlighted in bold; variables that are not statistically significant but feature in the best model are also presented.
Table 3.4 Model selection to estimate factors influencing each trophic level for the English Channel. Only the best supported models are shown. AICc weight: Akaike‟s Information Criteria (corrected) weights, values range from 0 to 1, high values indicating strong support for a given predictor; k: number of parameters in the model; R2: adjusted coefficient. Significant relationships are highlighted in bold; variables that are not statistically significant but feature in the best model are also presented.
CHAPTER 4 Figure 4.1 Map of the Celtic Sea region identified from the ICES divisions VIIf-g. Table 4.1 Parameterization of the Ecopath model for the Celtic Sea. B: Biomass; P/B: Production/Biomass
ratio
(instantaneous
rate
of
total
mortality);
Q/B:
Consumption/Biomass ratio (consumption represents the intake of food by a group); EE:
viii
Ecotrophic Efficiency express as fraction of the production of i that is consumed within the system, exported or harvested. Figure 4.2 The Celtic Sea Ecopath model in terms of relative biomass (size of circles) and its major energy flows within Functional Groups (FGs). The horizontal axis of symmetry of each box is aligned with the trophic level of this box. The value of a trophic level is a fractional because it depends on the diet composition of this group and on the trophic levels of its preys (Christensen and Pauly, 1993). Note that TL for Seabird surface feeders and Gulls the values were under estimated by Ecopath (respectively 3.82 and 2.46).
Figure 4.3 The Mixed Trophic Impacts Analysis in the Celtic Sea ecosystem. Seabird (impacted groups) response to the increase (by 10%) of other groups (impacting groups) biomass in the system. Positive impacts are shown above the base line in black (above the line) and negative impacts below in grey (below the line).
Figure 4.4 Trophic niche overlap index for seabird groups in the Celtic Sea calculated from the proportions of used resources in the diet matrix. Only groups with the highest values are shown. The index assumes values between 0 (no overlap) and 1 (total overlap).
CHAPTER 5 Table 5.1 Seabird and pelagic fish functional groups (FGs) used in the Ecosim scenarios. Seabird functional groups are adapted from JNCC, 2008. FGs structure in the model has previously described in Chapter 4 and Appendix 3. Figure 5.1 Simulated scenarios of changing fishery mortality (F) of pelagic fish and predicted effects on their relative biomass. The arrow refers to 2015 when the reform of Common Fishery Policy (CPF) will enter into force. (a) Application of Maximum Sustainable Yield (MSY): the largest yield (or catch) that can be taken from a species' stock over an indefinite period. (b) Simulation of no changes in fishery pattern over time. (c) Increase of F over time followed by no-take ban from 2025. Figure 5.2 Response of seabird groups to direct effects of simulated changes in pelagic fish fishery mortality (F) as outlined in Figure 1. (a) Application of Maximum Sustainable ix
Yield (MSY). (b) Simulation of no changes in fishery pattern over time. (c) Increase of F over time followed by no-take ban from 2025. Seabird offshore-surface feeders includes: northern gannet, black-legged kittiwake and northern fulmar; Seabird offshore divers includes: common guillemot, Atlantic puffin and razorbill; Gulls includes: lesser black backed gull, herring gull, great black backed gull, and black headed gull. Figure 5.3 Effects of cessation of discarding on seabird biomass in the Celtic Sea estimated from Ecosim model. (a) Seabird biomass change following a discard ban. (b) Seabird biomass change following a discard ban, but with model including increasing seabird foraging time. Biomass values for FGs were recorded at 2020. Figure 5.4 Seabirds response to climate forcing on Primary Production (PP). (a) Only changes to PP are applied while fishery is kept at constant value. (b) Both climate and fishery are modified; in particular fishery effort is reduced of 50% from 2015.
APPENDIX 1 Figure A1.1 Herring and sprat landings (kg/km2) from the Western and Celtic Sea Ground Fish Survey (WCGFS) (CEFAS). This trawl survey is designed to study the distribution, composition and abundance of all fish, commercial shellfish and cephalopod species in the Celtic Sea. Pearson‟s coefficient of correlation: 0.715, p value= 0.001. Table A1.1 Correlation matrix (Pearson‟s coefficient) between covariates. Significance is indicate as follow: pvalue< 0.001 ***, pvalue,< 0.01**, pvalue2mm); significant relationships are highlighted in bold, not significant variables included in the model are also presented. Table A1.4 Competing models for apex predators. AICc weight: Akaike‟s Information Criteria (corrected) weights, values range from 0 to 1, and high values indicate strong support for a given predictor; k: number of parameters in the model; R2: Adjusted coefficient. WNAO: winter North Atlantic Oscillation index; SNAO: spring North Atlantic Oscillation index; WSST: winter Sea Surface Temperature; 1lag-SSST: 1 year lagged spring Sea Surface Temperature; her 0-g: herring 0-group; her-1g: herring 1-group; Significant relationships are highlighted in bold, not significant variables included in the model are also presented.
APPENDIX 2 Table A2.1 Correlation matrix (Pearson‟s coefficient) between covariates for the Irish Sea. Significance is indicating as follow: p value
